Older adults with higher vitamin C levels showed stronger brain-network markers.
In recent years, the evolution of industrial sensing technologies has been pivotal to the advancement of intelligent manufacturing systems. Among various approaches, artificial neural networks (ANNs) have garnered immense attention due to their remarkable capability to model complex data relationships. However, applying these networks effectively in real-world industrial environments remains a formidable challenge, primarily due to the noisy, nonlinear, and often limited datasets encountered in such settings. Recognizing these constraints, a team of researchers led by Professor Lanxiang Sun at the Shenyang Institute of Automation, Chinese Academy of Sciences, has forged a transformative approach. Their groundbreaking solution, known as the Partial Least Squares-assisted Optimization Network (PLSaoNET), promises to redefine the application of neural networks in industrial sensing by embedding statistical interpretability into the heart of deep learning models.
Traditional neural networks typically commence training through random initialization of weights—a process that, while standard, lacks any guiding framework grounded in data characteristics. This randomness poses inherent limitations, especially when the amount of training data is limited or when the data exhibits high noise levels and nonlinear relationships, both common in industrial scenarios. Such conditions often result in poor generalization and unstable predictive outcomes, thereby hindering practical deployment. The innovation from Professor Sun’s team circumvents these issues through a statistical method known as Partial Least Squares (PLS), a classic tool renowned for its interpretability and robustness in linear regression tasks.
PLS works by extracting latent variables that capture the fundamental relationships between independent variables (input features) and dependent variables (output targets), offering a transparent statistical framework. While PLS excels in linear modeling, it lacks the capability to model complex nonlinear dynamics. The research team ingeniously leveraged the strengths of PLS by integrating its solutions as the initial conditions within a neural network architecture. Specifically, the number of PLS latent variables dictates the number of neurons in the network’s hidden layer, and the PLS-derived weight matrix initializes the neural connections. This initialization endows the neural network with a statistically meaningful starting point, effectively eliminating the guesswork inherent in random weight assignments.
Such a well-founded initialization transforms the learning process from an unfocused blind search into a systematic refinement of the PLS solution, integrating nonlinear modeling capabilities while preserving interpretability. This fusion of statistical rigor and deep learning adaptability addresses one of the fundamental bottlenecks in industrial data modeling, enabling the network to capture both the global trends identified by PLS and the nuanced nonlinear interactions vital for accurate predictions.
Moreover, industrial datasets frequently suffer from uneven label distributions, posing significant challenges during the training of neural networks using stochastic gradient descent. Fluctuations in loss values caused by biased mini-batch samples can lead to instability and degrade the overall learning process. To counter this, the researchers implemented a stratified sampling-based training strategy. In this approach, the entire dataset is divided into strata based on label value ranges, ensuring that each mini-batch during training samples from each stratum proportionally. This methodology smooths gradient updates, fortifying the network’s robustness against skewed data distributions and enhancing training stability substantially.
Professor Sun has emphasized that the primary motivation behind PLSaoNET is not to pursue architectural complexity but to advocate for models that produce stable, reliable outputs even when faced with small, noisy datasets that typify industrial environments. By embedding PLS’s statistical solutions into neural network training, the model’s learning trajectory gains a trustworthy, transparent foundation from which to build more complex representations.
To validate the efficacy of PLSaoNET, the research team conducted comprehensive experiments in two realistic industrial applications: the online monitoring of iron ore concentrate slurry grade using laser-induced breakdown spectroscopy (LIBS), and diesel fuel quality assessment with near-infrared (NIR) spectroscopy. Both cases involve complex spectral data with significant noise and nonlinear characteristics—ideal testbeds to showcase the robustness of PLSaoNET. Results consistently demonstrated superior prediction accuracy and generalization capabilities over traditional PLS regression and conventionally initialized backpropagation neural networks.
A particularly revealing insight emerged from the visualization and analysis of hidden layer weights in the LIBS slurry model. While the PLS model alone lacked precision, the dominant latent variables it extracted aligned clearly with known spectral emission lines from key mineral elements like silicon, magnesium, and iron, reflecting true physical relevance. In contrast, networks initialized randomly exhibited neuron weights that failed to selectively identify meaningful spectral features, resulting in noisy, uninterpretable representations. The PLSaoNET model, however, preserved the interpretability inherent to PLS by focusing attention on these dominant spectral features while simultaneously allowing fine-tuned retraining that accentuated secondary information related to elements like calcium and aluminum. This balance effectively suppresses overfitting and elevates model reliability.
Not content with theoretical validation alone, the team successfully deployed PLSaoNET within a live LIBS slurry analyzer at a mineral processing plant for real-time iron ore concentrate grade monitoring. This real-world application underscores PLSaoNET’s practical value in lowering the barriers to widespread adoption of artificial neural networks in industrial contexts. By anchoring the initial network parameters in statistically meaningful solutions and employing thoughtful training strategies, PLSaoNET represents a crucial step toward integrating cutting-edge AI models into complex, noise-prone industrial environments where data scarcity and variability have historically impeded progress.
Looking beyond its immediate applications, PLSaoNET delivers a broader paradigm shift, suggesting that harmonizing classical statistical techniques with modern deep learning can unlock new frontiers in intelligent industrial sensing. As manufacturing systems increasingly embrace digital transformation, methods like PLSaoNET that offer interpretability, stability, and adaptability will become indispensable tools for building resilient, data-driven infrastructures.
In conclusion, the work by Professor Lanxiang Sun and colleagues embodies a fusion of statistical principled approaches and adaptive machine learning techniques, producing an ANN model that addresses longstanding challenges in industrial data analysis. With its innovative PLS-based initialization and stratified sampling training, PLSaoNET not only enhances predictive performance but also paves the way for more interpretable, reliable AI systems in industrial manufacturing. As industries worldwide continue to seek intelligent sensing solutions that function robustly under real-world constraints, PLSaoNET stands out as a promising, practical breakthrough.
Subject of Research: Advanced artificial neural network design for improved industrial sensing under constraints of small, noisy, and nonlinear datasets.
Article Title: PLSaoNET: A Generalized ANN Model Under PLS Statistical Constraints for Industrial Sensing
News Publication Date: 5-May-2026
Web References:
http://dx.doi.org/10.1016/j.eng.2026.01.032
Image Credits: Lanxiang Sun
Industrial sensing, artificial neural networks, partial least squares, PLSaoNET, stratified sampling, nonlinear modeling, intelligent manufacturing, spectral analysis, laser-induced breakdown spectroscopy, near-infrared spectroscopy, deep learning initialization, model interpretability, industrial data analytics
A NASA investigation blamed millions of dollars in damage to one of its largest Deep Space Network (DSN) antennas on poor training and procedures.
The post Deep Space Network antenna mishap blamed on poor training and procedures appeared first on SpaceNews.
Ancient mycorrhizal fungal networks, consisting of a billion astronomical units’ worth of roots, have finally been mapped by scientists, providing insight into their connections, which are essential to climate regulation and plant sustenance.
Mycorrhizal fungi grow in an arbuscular formation, meaning they take on a branching tree-like structure of tubular cells called hyphae, which pull carbon from the atmosphere to aid in climate regulation.
The international team presented their work in a recent paper published in Science, which estimates the distribution and mass of these enormous fungal networks and provides an interactive visualization tool that offers crucial insight into threats to these systems, which are essential to Earth’s health.
The new research brings both encouraging and concerning findings, highlighting regions where fungal networks remain robust while also identifying areas under increasing threat.
Although previous studies have attempted to map the diversity of these fungi across the globe, this is the first effort to predict and visualize their density and biomass on a planetary scale. According to the team’s estimates, the network extends for approximately 110 quadrillion kilometers, with roughly 40% located beneath grasslands.
Using more than 16,000 soil cores collected worldwide, researchers assembled a massive fungal dataset. They then employed machine-learning models that combined data from multiple biomes to predict the density of arbuscular mycorrhizal fungal networks in unsampled regions. Robotic imaging of more than 300,000 living, laboratory-grown hyphae allowed the team to calibrate and validate their models.
“It is hard to overstate the importance and enormity of these fungi,” said lead author Dr. Justin Stewart, with the Society for the Protection of Underground Networks (SPUN). “There could be up to 10 meters (32 feet) of mycorrhizal network in just a teaspoon of soil.”
Because it transports carbon, water, and nutrients underground, scientists have likened the network to our planet’s circulatory system, with a flow rate of up to 120 micrometers per second. About 70% of Earth’s plant species thrive on a symbiotic relationship with the arbuscular mycorrhizal fungi network. The network can provide over 80% of a plant’s phosphorus and increase the root foraging area 100-fold.
The network subsists on carbon waste produced by plants, while providing them with water and nutrients in return. By pulling in roughly 4 billion tons of CO2 into the ground every year, mycorrhizal fungi mitigate an amount corresponding to roughly 11% of human emissions.
Unfortunately, grasslands, where the network thrives most, are among the least protected environments on Earth and are being rapidly converted to farmland. According to the study team, cropland has only about half the lower network density of regular land, which the researchers worry may negatively affect how soil stores carbon, cycles nutrients, and resists stress.
“It is important to remember that these maps are living documents. This is a first attempt at building a foundational model of Earth’s fungal networks infrastructure, but it is not complete,” Stewart recently told The Debrief. “There are other types of mycorrhizal fungi that still need to be mapped in more detail, and other fungal structures, such as spores, that are also critical to understanding how these organisms persist, move, and respond to environmental change. Likewise, we need more data in regions that have been poorly sampled.”
The team says the study represents a major advance for climate science, a field in which fungi have often been overlooked in discussions about conservation and resilience. Researchers hope the findings will encourage a more comprehensive understanding of the role fungi play in maintaining healthy ecosystems.
“A major goal now is to make these kinds of data useful for decision makers. Fungi are still rarely included explicitly in climate, biodiversity, and land-use agendas, even though they are central to how ecosystems move carbon and nutrients,” Stewart adds. “We want to help ‘course-correct’ by building a clearer picture of the underground systems that support life aboveground.”
The paper, “Global Density and Biomass of Arbuscular Mycorrhizal Fungal Networks,” appeared in Science on June 11, 2026.
Ryan Whalen covers science and technology for The Debrief. He holds an MA in History and a Master of Library and Information Science with a certificate in Data Science. He can be contacted at ryan@thedebrief.org, and follow him on Twitter @mdntwvlf.
NASA pushed its Deep Space Network beyond its limits during the Artemis I mission nearly four years ago. The global array of deep space communications antennas couldn't keep up with the routine demands of 40 robotic science missions and the extraordinary surge required by NASA's Orion space capsule as it flew around the Moon.
The experience in late 2022 reduced or delayed downlinks from several high-profile science missions, including the James Webb Space Telescope and Mars rovers, as the data-hungry Artemis I mission took priority on NASA's communications network. And that was before the first Artemis mission with astronauts onboard. When Artemis II launched April 1, NASA called upon the Deep Space Network (DSN) again to connect Mission Control to the Orion capsule as it soared more than a quarter of a million miles from Earth.
With a crew of four flying inside the spacecraft, the agency's appetite for data from Orion on Artemis II was even higher than it was on Artemis I. But at a little more than nine days, the Artemis II mission was shorter than the 25 days Artemis I spent in space, helping alleviate the communications overload. Artemis I also launched 10 small CubeSats into deep space, many of which required tracking and telecom services from the DSN. Artemis II carried fewer CubeSats.
© NASA/JPL-Caltech
NASA pushed its Deep Space Network beyond its limits during the Artemis I mission nearly four years ago. The global array of deep space communications antennas couldn't keep up with the routine demands of 40 robotic science missions and the extraordinary surge required by NASA's Orion space capsule as it flew around the Moon.
The experience in late 2022 reduced or delayed downlinks from several high-profile science missions, including the James Webb Space Telescope and Mars rovers, as the data-hungry Artemis I mission took priority on NASA's communications network. And that was before the first Artemis mission with astronauts onboard. When Artemis II launched April 1, NASA called upon the Deep Space Network (DSN) again to connect Mission Control to the Orion capsule as it soared more than a quarter of a million miles from Earth.
With a crew of four flying inside the spacecraft, the agency's appetite for data from Orion on Artemis II was even higher than it was on Artemis I. But at a little more than nine days, the Artemis II mission was shorter than the 25 days Artemis I spent in space, helping alleviate the communications overload. Artemis I also launched 10 small CubeSats into deep space, many of which required tracking and telecom services from the DSN. Artemis II carried fewer CubeSats.
© NASA/JPL-Caltech
5 min read
NASA has completed the investigation into the damage sustained last year at its 70-meter radio-frequency antenna, known as the Deep Space Station 14 (DSS-14), at the Goldstone Deep Space Communications Complex near Barstow, California. The agency has classified the event as a Type A mishap based on the total cost of damages. The antenna will remain offline to complete repairs and previously scheduled upgrades.
“NASA takes safety and any departure from established procedures seriously, and the investigation at Goldstone made clear that we must strengthen our processes. We are acting on the investigation’s findings,” said Joel Montalbano, acting associate administrator for NASA’s Space Operations Mission Directorate at the agency’s headquarters in Washington. “We will update and improve procedures, rebuild core in-house capabilities, and reinforce operational discipline across the Deep Space Network. NASA remains focused on learning from this and modernizing systems, so DSS-14 and the broader network are ready to support our ambitious future missions.”
On Sept. 16, 2025, the DSS‑14 antenna over‑rotated while actively tracking the Juno mission, placing excessive stress on cabling and associated structural supports. Water lines tied to the antenna’s fire‑suppression system also were damaged, causing significant flooding in the facility. There were no injuries.
NASA convened a Mishap Investigation Board, bringing together experts from across the agency to examine the technical, organizational, and cultural factors behind the incident. The board conducted on‑site inspections, interviews, and detailed reviews of technical documentation and operational logs from all three Deep Space Network sites. The board completed its final report in April and submitted it for agency concurrence.
The investigation issued findings and recommendations that emphasize training, technical rigor, operational procedures, system design, clear roles and responsibilities, and safety assurance. At the same time, teams already are applying lessons learned across all network sites to improve operational consistency. These steps will help bolster the network and reduce the risk of future mishaps.
In its final report, the board found the mishap primarily stemmed from software weaknesses, human error, and an undetected failure in the antenna’s hydraulic limit system. Investigators determined an electrical issue at the antenna the previous day caused the control system to misreport the antenna’s rotation state, an issue that went unnoticed and triggered multiple limit-stops during the Juno track on Sept. 16. While working to identify the limit-stop problem, operators performed several troubleshooting steps that inadvertently bypassed software and hardware safeguards, which ultimately led to the over-rotation incident. After flooding in the antenna base was observed, operators attempted to stow the antenna as a safety precaution, however, because the system had already passed the rotation limits, this action drove the antenna further into over‑rotation, causing additional damage.
Additionally, the investigation found the antenna’s hydraulic limit system, its final mechanical safeguard, was inoperable on Sept. 16 after being damaged in an undocumented prior incident. The system also had not been adequately tested for an undetermined period of time.
Investigators also concluded workplace culture pressured operators to work as expeditiously as possible, often stretching beyond their usual roles, expertise, and training, to keep the antenna operating. The board states the cultural conditions observed at Goldstone were not present at the network’s other sites, where roles and responsibilities are followed more consistently. Other contributing factors outlined in the report include inadequate procedures, reliance on undocumented practices and tacit knowledge, and gaps in the antenna’s control logic. NASA will accept this as the final report.
The agency estimates repairs will cost between $4.1 and $4.6 million, with a final figure to be determined after the antenna’s systems are fully assessed. The antenna will remain offline as it enters its previously scheduled extended maintenance and upgrade period, originally set to begin in August and expected to be completed by October 2028. These upgrades are part of broader network improvements essential to supporting future exploration and science missions, as well as enhancing the nation’s planetary defense capabilities.
“We are committed to learning everything we can from this incident, and we’ve already begun putting those lessons into practice,” said Kevin Coggins, deputy associate administrator for NASA’s SCaN (Space Communications and Navigation) Program at the agency’s headquarters. “Our teams are working to strengthen and standardize processes and training across all three network sites to ensure it remains resilient, consistent, and ready to support the next generation of missions. Every challenge is an opportunity to improve, and this is no exception.”
The Deep Space Network continues to provide full coverage for more than 40 missions despite the DSS‑14 incident. The network’s 13 other antennas, located at complexes in California, Australia, and Spain, are supporting all tracking needs without interruption. A dedicated scheduling team allocates antenna time across the network to meet each mission’s science and data‑return objectives. The team also maintains continuous coverage when an antenna goes offline for maintenance or an unexpected outage.
To view the report, which includes redactions to protect proprietary and privacy-sensitive material, visit:
SpaceX has won a lucrative contract to provide the US military with a means of distributing space-based sensing and targeting data, forming the "backbone" of a rearchitected network after separate Pentagon initiatives stalled, officials announced Tuesday.
Space Systems Command, the Space Force's primary procurement and acquisition center, announced the $2.29 billion firm-fixed-price agreement, confirming long-simmering reports that the Pentagon was likely to tap SpaceX for a new communications network in low-Earth orbit. SpaceX's selection for the Space Data Network (SDN) Backbone contract "accelerates the delivery of a resilient, high-speed communications network in space," Space Systems Command said in a statement.
The network will be based on technology originally developed for SpaceX's Starlink global Internet constellation. SpaceX already builds and launches specially designed satellites, called Starshield, for military applications. The SDN Backbone network in low-Earth orbit (LEO) will presumably use the Starshield platform.
© US Space Force/Gwendolyn Kurzen
Login